More on Crist's Steep Climb

I've gotten a fair number of emails and calls today on my blog post saying that Crist won't win, mostly from Democrats who hope that I am wrong because they see Crist as the best chance to beat Rubio.  Several of them wanted to know how I saw Meek winning, which if you read the post, isn't the point.  Even the Weekly Standard had a take.  Trust me, I didn't see that one coming.
 
Nonetheless, the chatter made it clear the Crist theory needed more explanation.  So let me delve a little more into the numbers.
 
Going back to the basic math to win for Crist:  win roughly 1/3 of the partisan vote and 50% of the NPA.   First, there is one big assumption there, that Crist will get 50% of the NPA (no party affiliation) vote.  Some polls suggest he could get more, but many more show him under 50 with NPA voters. Quite frankly, if he doesn't get 50% of the NPA vote, the whole conversation is purely academic.  

Nonetheless, under that assumption and a model assuming that a few more Dems vote on election day than republicans, which given the Dems six point registration advantage is likely, Crist wins by a few points.  But here is problem one:  Who really thinks Crist will get 33% of the GOP vote?  Not this observer, nor any Republican hack who I trust. 
 
Looking more into the math, and again, assuming that Crist gets 50% of the NPA vote, which in a three way race is far from a sure thing, here is what he would need to get from R's and D's to get a plurality of vote.  This is all built off a 42-41-17 R-D-NPA election day model:
 
If 33% of GOP vote for him, he needs about the same in Dem vote.
If 25% of GOP votes for him, he needs 41% of the Dem vote (by less than a point).
If 20% of GOP votes for him, he needs 50% of the Dem vote to win (by four-tenths).
If 18% of GOP votes for him, he needs 53% of the Dem vote to win (by two-tenths).
If 15% of GOP votes for him, he needs 59% of Dem vote to win (by three-tenths).

Does anyone really think any of these scenarios are likely?
 
Further, all of this assumes that in this kind of race, Meek would get 5% of Republicans.  In fairness, I think there is a better chance that Rubio gets 5% of Dems than Meek gets 5% of Republicans.  It also assumes that NPA breaks evenly between Meek and Rubio -- which is a really big if.  
                                         
Could Democrats leave en masse for Crist?  Sure, on an excel spreadsheet they could.  

But will they?  Let's play that one out.
 
First of all, if we assume (and I think this is a fair assumption since Davis got 81%), that Meek will get at least 75% of the African American/Caribbean vote, which makes up roughly 13% of the likely total statewide turnout, that alone gets him to roughly 10 points statewide, virtually all out of the Democratic column.  

As a baseline to get to a win, if we work off the alerady stretch scenario that Crist gets 25% of the GOP vote, Crist needs to get 41% of the Democratic vote. But if Meek get 75% of the African American/Caribbean vote, Crist would need 55-56% of the remaining white and Hispanic Democratic vote. Is that going to happen?  Do I think Garrard leads the Jaguars to a Super Bowl win?  

That would be no and no. 
 
What can Crist do to increase his Democratic support?  He could announce he is caucusing with the Democrats.  That may win him some votes, but some Democrats will see it as pandering and stick with Meek.  Who else will see it as pandering:  Republicans and Independents.  If Crist says he is with the Democrats, there is a zero percent chance he gets to an alrady unlikely floor of 25% of Republicans. 
 
In addition, one other factor comes into play:  the hardening of partisanship.  Elections always narrow because partisans come home closer to election day, which is why in Florida, winning the NPA vote in a two way statewide or close district-level race is vital for winning.  In a two-way race, that pushes both candidates to the middle, but in a three-way race, it hurts the one in the middle.  If Crist doesn't get close to a third of the total partisan vote, he will lose, plain and simple.

There's another scenario where he never even comes close.  For example, if he only gets 25% of the two party vote, he would need to get 70% of the NPA vote to get to 34%, which probably won't be enough to win.  Even at 70% of the NPA vote, he'd still needs 29-30% of the two party vote to get to a reasonable win number. 

Meek's path is also rough - arguably rougher than Crist.   Rubio is much more likely to consolidate Republicans --even with Crist in the race. Meek really can't consolidate Dems with Crist in the race -- and Meek needs a good showing with NPA voters, which will be impossible with Crist in the race.  
 
But Crist's path is uncertain at best -- and it is clear his team is struggling to figre out what it is.  Confident candidates stay on message, while less confident candidates scramble around to find solid ground.  Is there any question that Crist is scrambling?
 
Now you ask, can Kendrick Meek win?
 
Go back to the original Election Day turnout scenario, with Crist winning NPA voters with 50% and Meek/Rubio split the rest.  They each get 80% of their party vote, because Democrats come home understanding that Meek is the best chance.  What happens then?  Meek wins 40-39-21.  Sure he can win.  Will he win?  That is another question.

People keep asking "But isn't a better strategy for Democrats to vote for Crist?"   First, I don't think voters are that 'processey.' Secondly, the 'rally around Crist' theory also assumes that the several million or so Democrats likely to vote in November are monolithic.  Trust me, they aren't.

Maybe if Jeff Greene had won the primary, the math would be different, but I'd still be skeptical. But Greene didn't, by a long shot.

Simply, I don't see as many as half of Democrats statewide abandoning Kendrick Meek and voting for Charlie Crist, especially with both President Clinton and Obama supporting him, institutional Democrats rallying around him and the vast majority of Democratic electeds remaining firmly in his camp, which is why the time to come to embrace reality and for Democrats to help Meek. 

Otherwise, Marco will remain firmly in the driver's seat, which is where he sits today.

Comments

[joan mckniff]

Your analysis, logic and math are right on the mark. The handful of Dems I don't understand are those who tell me that they are concerned Meek might not win in Nov so they are not working for his election now.

[Tally]

A better strategy for Crist for getting Dem votes is endorsing Alex Sink for governor and forcing Rubio to endorse Rick Scott. I think the 20%R and 50%D scenario is quite doable for Crist under this scenario.

Meek beat Greene with help from Pinellas and Hillsborough counties. If those voters go to home town guy Crist, it would be easy for him to get 50%D.

The only way Meek might succeed is if he is able to energize young voters in larger numbers. But he is going to have to get outside his comfort zone and become a much better candidate to do that.